keraflow.backend.tensorflow_backend.TensorflowBackend Class Reference

Inheritance diagram for keraflow.backend.tensorflow_backend.TensorflowBackend:

Public Member Functions
def	shape
	Returns the symbolic shape of a tensor.
def	int_shape
	Returns the shape of a tensor as a tuple of integers or None entries. More...
def	ndim
	Returns the number of axes in a tensor, as an integer.
def	dtype
	Returns the dtype of a tensor, as a string.
def	eval
	Evaluates the value of a tensor. More...
def	cast
	Casts a tensor to a different dtype.
def	set_value
	Sets the value of a tensor variable, from a Numpy array.
def	switch
	Switches between two operations depending on a scalar value (int or bool). More...
def	function
	Instantiates a Keraflow function. More...
def	gradients
	Returns the gradients of variables (list of tensor variables) with regard to loss.
def	zeros
	Instantiates an all-zeros tensor variable.
def	ones
	Instantiates an all-ones tensor variable.
def	eye
	Instantiate an identity matrix.
def	zeros_like
	Instantiates an all-zeros tensor of the same shape as another tensor.
def	ones_like
	Instantiates an all-ones tensor of the same shape as another tensor.
def	dot
	numpy.dot on tensors
def	gather
	Retrieves the vectors of indices indices in the 2D tensor reference. More...
def	prod
	Multiplies the values in a tensor, alongside the specified axis.
def	mean
	Mean of a tensor, alongside the specificied axis.
def	std
	Standard deviation of a tensor, alongside the specificied axis.
def	any
	Bitwise reduction (logical OR). More...
def	argmax
	Returns the index of the maximum value along a tensor axis.
def	argmin
	Returns the index of the minimum value along a tensor axis.
def	square
	Element-wise square.
def	abs
	Element-wise absolute value.
def	sqrt
	Element-wise square root.
def	exp
	Element-wise exponential.
def	log
	Element-wise log.
def	round
	Element-wise rounding to the closest integer.
def	sign
	Element-wise sign.
def	pow
	Element-wise exponentiation.
def	clip
	Element-wise value clipping.
def	equal
	Element-wise equality between two tensors. More...
def	not_equal
	Element-wise inequality between two tensors. More...
def	maximum
	Element-wise maximum of two tensors.
def	minimum
	Element-wise minimum of two tensors.
def	sin
	Computes sin of x element-wise.
def	cos
	Computes cos of x element-wise.
def	concatenate
	Concantes a list of tensors alongside the specified axis.
def	reshape
	Reshapes a tensor to the specified shape.
def	transpose
	Permutes axes in a tensor. More...
def	repeat
	Repeats the elements of a tensor along an axis, like np.repeat. More...
def	expand_dims
	Adds a 1-sized dimension at index "axis".
def	squeeze
	Removes a 1-dimension from the tensor at index "axis".
def	sigmoid
	Element-wise sigmoid.
def	hard_sigmoid
	Segment-wise linear approximation of sigmoid. More...
def	tanh
	Element-wise tanh.
def	relu
	Rectified linear unit. More...
def	softmax
	Softmax of a tensor.
def	softplus
	Softplus of a tensor.
def	dropout
	Sets entries in x to zero at random, while scaling the entire tensor. More...

Member Function Documentation

def keraflow.backend.tensorflow_backend.TensorflowBackend.any	(		self,
			x,
			axis = None,
			keepdims = False
	)

Bitwise reduction (logical OR).

Returns an uint8 tensor (0s and 1s).

def keraflow.backend.tensorflow_backend.TensorflowBackend.dropout	(		self,
			x,
			drop_rate,
			noise_shape = None
	)

Sets entries in x to zero at random, while scaling the entire tensor.

Parameters

x	tensor
drop_rate	fraction of the entries in the tensor that will be set to 0.
noise_shape	shape for randomly generated keep/drop flags, must be broadcastable to the shape of x

def keraflow.backend.tensorflow_backend.TensorflowBackend.equal	(		self,
			x,
			y
	)

Element-wise equality between two tensors.

Returns a bool tensor.

def keraflow.backend.tensorflow_backend.TensorflowBackend.eval	(		self,
			x
	)

Evaluates the value of a tensor.

Returns a Numpy array.

def keraflow.backend.tensorflow_backend.TensorflowBackend.function	(		self,
			inputs,
			outputs,
			updates = [],
			kwargs
	)

Instantiates a Keraflow function.

Arguments

inputs: list of placeholder/variable tensors. outputs: list of output tensors. updates: list of update tuples (old_tensor, new_tensor).

def keraflow.backend.tensorflow_backend.TensorflowBackend.gather	(		self,
			reference,
			indices
	)

Retrieves the vectors of indices indices in the 2D tensor reference.

Parameters

reference	a 2D tensor.
indices	2D int tensor or list.

Returns

3D tensor.

def keraflow.backend.tensorflow_backend.TensorflowBackend.hard_sigmoid	(		self,
			x
	)

Segment-wise linear approximation of sigmoid.

Faster than sigmoid.

def keraflow.backend.tensorflow_backend.TensorflowBackend.int_shape	(		self,
			x
	)

Returns the shape of a tensor as a tuple of integers or None entries.

Note that this function only works with TensorFlow.

def keraflow.backend.tensorflow_backend.TensorflowBackend.not_equal	(		self,
			x,
			y
	)

Element-wise inequality between two tensors.

Returns a bool tensor.

def keraflow.backend.tensorflow_backend.TensorflowBackend.relu	(		self,
			x,
			alpha = 0.,
			max_value = None
	)

Rectified linear unit.

Arguments

alpha: slope of negative section. max_value: saturation threshold.

def keraflow.backend.tensorflow_backend.TensorflowBackend.repeat	(		self,
			x,
			rep,
			axis
	)

Repeats the elements of a tensor along an axis, like np.repeat.

If x has shape (s1, s2, s3) and axis=1, the output will have shape (s1, s2 * rep, s3)

def keraflow.backend.tensorflow_backend.TensorflowBackend.switch	(		self,
			condition,
			then_expression,
			else_expression
	)

Switches between two operations depending on a scalar value (int or bool).

Note that both then_expression and else_expression should be symbolic tensors of the same shape.

Arguments

condition: scalar tensor. then_expression: TensorFlow operation. else_expression: TensorFlow operation.

def keraflow.backend.tensorflow_backend.TensorflowBackend.transpose	(		self,
			x,
			dims
	)

Permutes axes in a tensor.

Arguments

dims: should be a tuple of dimension indices, e.g. (0, 2, 1).